On Wed, Jul 09, 2008 at 02:20:55PM -0700, H.J. Lu wrote: > On Wed, Jul 9, 2008 at 1:59 PM, Joe Buck <[EMAIL PROTECTED]> wrote: > > On Wed, Jul 09, 2008 at 01:52:27PM -0700, Steve Kargl wrote: > >> > >> Do you have an application that shows that the current > >> implementation of chmod() is a bottleneck? > > > > Yes, this might be a case of "if it ain't broke, don't fix it". > > > > It is broken:
Do you have an application that shows that it is broken? > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36755 I've read the PR. > The implementation in libgfortran is incorrect. It doesn't > handle signal nor check return from wait to match pid. In that chmod() is not defined by the Fortran Standard, how can you infer that libgfortran's implementation is incorrect? In fact, the authorative documentation would be the g77 manual (because gfortran provides chmod solely for backwards compatibility). It clear states: Note that this currently works by actually invoking /bin/chmod (or the chmod found when the library was configured) and so might fail in some circumstances and will, anyway, be slow. -- Steve
